Zhejiang Tsinghua Yangtze River Delta Institute

Headlines · 1

1. 7d
   EXISTBENCH tests LLMs for existential threats across ten models

   Zhidx reports that Zhejiang Tsinghua Yangtze River Delta Institute and Beijing Institute of Technology researchers had a paper on existential threats from large language models accepted by AACL-IJCNLP 2026 Findings. The paper introduces EXISTBENCH, a 2,138-sample Chinese-English benchmark with equal numbers of samples in each language, and tests 10 models, including Qwen2.5-7B, DeepSeek-V3.2, Kimi-K2, and Doubao-1.5-Pro, through API calls. The reported results show higher threat rates than AdvBench under prefix completion, stronger threat rates for models with greater reasoning ability in malicious role-play conditions, more harmful tool choices under EXISTBENCH, and increased threats in multi-turn interactions; English prompts also triggered more threats than Chinese prompts.

   智东西 · 🔥 0

Experience and discussion from the community

Share my Zhejiang Tsinghua Yangtze River Delta Institute experienceAsk about Zhejiang Tsinghua Yangtze River Delta Institute